Growth kinetics of platinum nanocrystals prepared by two different methods: Role of the surface

Journal of Colloid and Interface Science 365 (2012) 117–121

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Growth kinetics of platinum nanocrystals prepared by two different methods: Role of the surface Neenu Varghese, C.N.R. Rao ⇑ Chemistry and Physics of Materials Unit, New Chemistry Unit and CSIR Centre of Excellence in Chemistry, Jawaharlal Nehru Centre for Advanced Scientific Research, Jakkur P.O., Bangalore 560 064, India

a r t i c l e

i n f o

Article history: Received 2 July 2011 Accepted 3 September 2011 Available online 10 September 2011 Keywords: Platinum nanocrystals Diffusion-limited growth Surface reaction-limited growth Small angle X-ray scattering Transmission electron microscopy

a b s t r a c t In order to examine the applicability of the diffusion-limited Ostwald ripening model to the growth kinetics of nanocrystals, platinum nanocrystals prepared by two different methods have been investigated by a combined use of small-angle X-ray scattering (SAXS) and transmission electron microscopy (TEM). One of the methods of synthesis involved the reduction of chloroplatinic acid by sodium citrate while in the other method reduction was carried out in the presence of polyvinylpyrrolidone (PVP) as a capping agent. The growth of platinum nanocrystals prepared by citrate reduction in the absence of any capping agent follows a Ostwald ripening growth with a D3 dependence. In the presence of PVP, the growth of platinum nanocrystals does not completely follow the Ostwald ripening model, making it necessary to include a surface reaction term in the growth equation. Thus, the growth of platinum nanocrystals in the presence of PVP has contributions both from diffusion and surface reaction, exhibiting a D3 + D2 type behavior. Ó 2011 Elsevier Inc. All rights reserved.

1. Introduction Growth kinetics of nanocrystals is of interest due to its importance in understanding the synthesis of nanostructures of different shapes and sizes [1,2]. The most popular mechanism employed to explain the growth kinetics of nanocrystals is the diffusion-limited Ostwald ripening process following the Lifshitz–Slyozov–Wagner (LSW) theory [3,4]. Studies of the growth of ZnO and TiO2 nanocrystals by Searson and coworkers [5–7] have revealed Ostwald ripening to be the dominant mechanism. The study of Kondow et al. on platinum nanoparticles generated by laser ablation showed that Ostwald ripening was the growth mechanism in the absence of any surfactant [8]. Simonsen et al. [9] employed TEM for a in situ study of Pt nanoparticle sintering on a Al2O3 support and observed the growth of the particles to follow the Ostwald ripening process. The study of Peng and Peng [10,11] on the growth kinetics of CdSe nanorods by UV–vis spectroscopy and TEM indicated that the diffusion-controlled model applied when the monomer concentration was high. The growth of CdSe and InAs nanocrystals studied by Alivisatos and coworkers [12] showed the occurrence of focusing and defocusing in size distribution similar to that expected in Ostwald ripening. A diffusion-limited mechanism was found to occur in the last stage of the growth in the case of CdSe nanocrystals by Qu et al. [13]. A detailed study of the growth mechanism of

⇑ Corresponding author. Fax: +91 80 22082760. E-mail address: [email protected] (C.N.R. Rao). 0021-9797/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2011.09.005

uncapped ZnO nanorods by Biswas et al. [14] based on SAXS and TEM showed a diffusion-controlled growth of ZnO nanorods. There are a few reports in the literature where the growth kinetics of nanocrystals are found to deviate from the simple diffusion limited Ostwald ripening model. Thus, Seshadri et al. [15] proposed the growth of gold nanoparticles to be essentially stochastic wherein the nucleation and growth steps are well separated. A recent study of the nucleation and growth of gold nanocrystals by SAXS and UV–vis spectroscopy over very short time scale suggests a surface reaction-limited growth of nanocrystals in the presence of an alkanoic acid [16]. A recent study of the slow growth kinetics of gold nanocrystals also showed a sigmoidal growth law to operate rather than Ostwald ripening [17]. These results are contrasted with those of El-Sayed and coworkers [18] who found diffusion-controlled growth of small gold nanoclusters based on a TEM investigation. Theoretical investigations suggest that the growth of nanocrystals could be controlled by the diffusion of particles as well as the reaction at the surface [19]. Thus, Viswanatha et al. [20] observed the growth of the ZnO nanocrystals in water to follow a growth mechanism intermediate between diffusion-control and surface reaction-control. They report that the growth kinetics of the ZnO nanocrystals to deviate from diffusion-controlled Ostwald ripening even in the absence of any capping agent [21]. Contributions from both diffusion and surface reactions have been reported in the case of CdSe and CdS nanocrystals [22]. The above discussion clearly shows how Ostwald ripening as well as other growth models have been reported for both metallic and semiconducting nanomaterials in solution in the literature. It

N. Varghese, C.N.R. Rao / Journal of Colloid and Interface Science 365 (2012) 117–121

was also seen that surface plays an important role in determining the mechanism. It was of interest to see why Ostwald ripening is not observed as the growth mechanism in the case of nanoparticles and how the surface plays a role when there is a deviation from this mechanism. With this purpose, we have studied the growth kinetics of platinum nanocrystals prepared by two different procedures. In the first synthesis, the platinum nanoparticles were prepared by the reduction of hexachloroplatinate solutions by sodium citrate [23]. In the second synthesis, chloroplatinic acid was reduced in ethylene glycol in the presence of PVP as the capping agent [24]. We have employed SAXS along with TEM to investigate the growth kinetics.

2. Experimental Platinum nanoparticles were prepared by the reduction of chloroplatinic acid by sodium citrate using the procedure reported in the literature [23]. In a typical reaction, 4 mL of 19.3 mM H2PtCl6 solution was added to 110 mL of distilled water and was brought to boiling temperature. To the boiling solution, 10 mL of aqueous solution containing 0.15 g of sodium citrate was added and the mixture was refluxed. Small aliquots were collected from the reaction mixture after different reaction times and were cooled in an ice bath to stop the growth. These solutions were used for further measurements. In the second procedure, platinum nanocrystals were prepared by the reduction in ethylene glycol as reported in the literature [24]. In a typical synthesis, 1 mL of 80 mM H2PtCl6 solution in ethylene glycol was rapidly added to 7 mL ethylene glycol which was held at 160 °C that contained both NaNO3 and PVP. The concentration of NaNO3 and PVP were 33 mM and 30 mM respectively. The mixture was stirred and maintained at 160 °C for the growth. Small aliquots were collected from the reaction mixture after different reaction times and were cooled in an ice bath to stop the growth. Acetone was added to these solutions to precipitate the nanocrystals. The precipitate was collected by centrifugation, dried and redissolved in ethanol for further measurements. The average diameter of the platinum nanocrystals could be readily obtained by SAXS [25–27]. We performed SAXS experiment with a Bruker-AXS NanoSTAR instrument modified and optimized for solution scattering. The instrument is equipped with a X-ray tube (Cu Ka radiation, operated at 45 kV/35 mA), cross-coupled Göbel mirrors, three-pinhole collimation, evacuated beam path, and a 2D gas-detector (HI-STAR) [27]. The accessible scattering range of the instrument can be varied by selecting different distances between the sample and the detector of 42.2 cm and 66.2 cm. The modulus of the scattering vector is q ¼ 4pSinh=k, where h is scattering angle and k is X-ray wave length. We recorded the SAXS data in the q = 0.007–0.21 Å1 range, i.e., 2h = 0.1–3°. Solutions of the platinum nanocrystals in water or ethanol (approximate 0.1 w/v% concentration) obtained after lapse of different reaction times were used for SAXS measurements. The solutions of nanocrystals were taken in quartz capillaries (diameter of about 2 mm) for the measurements with an exposure time of 30 min in order to obtain good signal-to-noise ratios. A capillary filled with only water or ethanol was used for background correction. The concentration of the nanocrystals was sufficiently low to neglect interparticle interference effects. The experimental SAXS data were fitted by Bruker-AXS DIFFRACplus NANOFIT software by using solid sphere model. The form factor of the spheres used in this software is that due to Rayleigh [28]. The assumption of the spherical model was verified by carrying out TEM investigations at several points of the growth process. The solutions of platinum nanocrystals obtained after different reaction periods were taken on holey carbon-coated Cu grids for

TEM investigations with a JEOL (JEM3010) microscope operating with an accelerating voltage of 300 kV. The diameter distributions were obtained from the magnified micrograph by using DigitalMicrograph 3.4 software. 3. Results and discussion Results of the SAXS measurements on the growth of platinum nanocrystals prepared by the citrate reduction are shown in Fig. 1 with logarithmic plots of intensity vs. scattering vector for different times of growth. The changes observed in the SAXS patterns indicate that the nanocrystals grow as a function of time. In order to estimate the average diameter and diameter distribution of platinum nanocrystals, the experimental SAXS data is fitted to the spherical model mentioned in Section 2. The scattering contrast for X-rays is given by the electron density difference between the particle and the solvent. Since platinum has much higher contrast than the solvent, only the platinum particles are considered for analysis. In the case of a dilute assembly of spherical particles, neglecting particle interaction, the scattering intensity is given by [16],

IðqÞ /

Z

f ðrÞVðrÞ2 Pðq; rÞdr

ð1Þ

Here, V(r) and P(q, r) are the volume and form factors respectively of a sphere of radius r. The form factor of the sphere is given by [28],

Pðq; rÞ ¼

" #2 3fsinðqrÞ  qr cosðqrÞg

ð2Þ

ðqrÞ3

In Eq. (2), q is the scattering vector (q ¼ 4pSinh=k, where h is the scattering angle and k is X-ray wave length), and r the radius of the sphere. The particle size distribution is obtained using the Gaussian distribution, f(r), 2 1 2 f ðrÞ ¼ pffiffiffiffiffiffiffi e½ðrr0 Þ =2r  r 2p

ð3Þ

Least square refinement yields two parameters, the radius, r and the standard deviation, r. The solid curves in Fig. 1 are fits of the

t (min) 440 330

log (I)

118

240 210 150 75

40 -2.0

-1.6

-1.2

-0.8

log (q) Fig. 1. SAXS data for the growth of platinum nanocrystals prepared by the reduction with sodium citrate after different reaction times. Solid lines are the spherical model fits to the experimental data.

119

N. Varghese, C.N.R. Rao / Journal of Colloid and Interface Science 365 (2012) 117–121

2.8

D (nm)

2.4

SAXS TEM 3 D fit

2.0

1.6

1.2 0

100

200

300

400

500

t (min)

Fig. 2. Time evolution of diameter distribution of the platinum nanocrystals prepared by reduction with sodium citrate. The distributions are obtained from the spherical model fits to the experimental SAXS data, recorded after different reaction times.

experimental SAXS data to the sphere model. The fits with the experimental pattern are quite good, yielding the average diameter as well as the diameter distributions for different growth times. We have estimated the diameter, D, of the platinum nanocrystals after different periods of growth from the SAXS data, and have shown the time evolution of the diameter distribution in Fig. 2. The width of the diameter distribution increases first and then decreases for different times of growth. Such focusing and defocusing of the diameter and length distributions [12,14] occurs when the monomer (atomic or molecular species involved in the growth process) concentration gets depleted due to the faster growth of the nanocrystals and the smaller nanocrystals shrink as the bigger ones grow. The size distribution, therefore, becomes broader. Dissolution of the small nanocrystals again enriches the monomer concentration in the solution, with the growth of the bigger nanocrystals continuing through the diffusion of the monomer from solution to the nanocrystals surface resulting in the focusing of the diameter distribution. Focusing and defocusing of the diameter distribution is thus dependent on the variation of the monomer concentration in the solution and indicates that the growth of platinum nanocrystals occurs mainly by the diffusion of monomer from solution to surface of the nanocrystals.

0.80.8 0.6

(b)

0.8 0.8 0.6 0.6

NF

0.6

1.0 1.0

NF

NF

Fig. 5. SAXS data for the growth of platinum nanocrystals prepared by ethylene glycol reduction in the presence of PVP after different times of the reaction. Solid lines are the spherical model fits to the experimental data.

(a)

1.01.0

NF

Fig. 4. Time evolution of the average diameter (D) of the platinum nanocrystals obtained from SAXS (filled circles). TEM data are shown by open circles. Solid curve represents fit of the data to the diffusion-limited Ostwald ripening growth model.

0.40.4

0.4 0.4 0.2 0.2

0.20.2 0.00.0 1.2

1.6

D /nm2.0 D /nm

10 nm

2.4

0.0 0.0 2.0

2.4

2.8

3.2 D /nm

3.6

4.0

D /nm

10 nm

Fig. 3. TEM images of platinum nanocrystals prepared by the reduction with sodium citrate after (a) 40 and (b) 440 min of growth. Insets show the diameter distributions. NF is the normalized frequency.

120

N. Varghese, C.N.R. Rao / Journal of Colloid and Interface Science 365 (2012) 117–121

(a)

1.0 1.0

0.6 0.6

0.4 0.4

0.4 0.4

0.2 0.2

0.2 0.2

0.0 0.0 2.0

(b)

0.8 0.8

NF

0.6 0.6

1.0 1.0

NF

NF

NF

0.8 0.8

2.5

3.0

3.5

D (nm)

4.0

0.0 0.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5

4.5

D (nm)

D (nm)

20 nm

20 nm

Fig. 6. TEM images of platinum nanocrystals prepared by ethylene glycol reduction in the presence of PVP after (a) 3 and (b) 110 min of growth. Insets show the diameter distributions. NF is the normalized frequency.

D3  D30 ¼ Kt

ð4Þ

where D is the average diameter at time t and D0 average initial diameter of the nanocrystals. The rate constant K is given by 2 K ¼ 8cdV m C 1 =9RT, where d is the diffusion constant at temperature T, Vm the molar volume, c the surface energy, and Ca the equilibrium concentration at a flat surface. We fitted the D(t) data of the platinum nanocrystals obtained from SAXS and TEM to the Ostwald ripening model by Eq. (4). The goodness of fits with R2 = 0.99 over the entire range of experimental data by the Ostwald ripening model is shown by solid curve in Fig. 4. We have studied the effect of capping agent on the growth kinetics of platinum nanocrystals by employing the second procedure for the synthesis. In this procedure, particles were prepared by ethylene glycol reduction of the platinum salt in the presence of PVP as the capping agent. Fig. 5 shows typical intensity vs. scattering vector plots in the logarithmic scale for different times obtained from SAXS measurements. We have fitted the experimental SAXS data to the spherical model and also carried out TEM measurements on the platinum nanocrystals at a few points during the growth. We show typical TEM images of the nanocrystals after 3 and 110 min of growth in Fig. 6. The diameter distributions obtained from TEM and SAXS show good agreement with the diameter varying from 3.2 to 4.8 nm over 120 min.

4.8

TEM 4.4

SAXS 3

2

D +D

D (nm)

We have carried out TEM measurements during the growth of the platinum nanocrystals. In Fig. 3, we show typical TEM images of platinum nanocrystals prepared after 40 and 440 min of growth to illustrate how the size of the platinum nanocrystals increases with increasing time. We have estimated the average diameter and the diameter distribution of platinum nanocrystals at different stages of the growth from the TEM images. The diameter distributions obtained from TEM and SAXS show good agreement with each other. Fig. 4 shows the time evolution of the average diameter of the platinum nanocrystals obtained from SAXS (filled squares) and TEM (open circles). The diameters obtained from the two techniques are comparable, the values varying between 1.2 and 2.9 nm over a period of 440 min. Although some evidence for the diffusion-limited growth of platinum nanocrystals is provided by the observation of the focusing and defocusing diameter distribution, a better insight is obtained from the time evolution of the average diameter of nanocrystals. If the diffusion-limited Ostwald ripening according to LSW theory [3,4] were the sole contributor for the growth, the rate law would be given by [2],

Ostwald

4.0

3.6

3.2

0

20

40

60

80

100

120

t (min) Fig. 7. Time evolution of the average diameter (D) of the platinum nanocrystals prepared by ethylene glycol reduction in the presence of PVP obtained from SAXS (filled squares). TEM data are shown by open circles. Dotted curve is the Ostwald ripening model fit to the experimental data. Solid curve represents the combined surface and diffusion growth model fits to the experimental data.

We have plotted the average diameter obtained from SAXS (filled squares) and TEM (open circles) of platinum nanocrystals against time in Fig. 7. We have tried to fit the D(t) data to the Ostwald ripening model [Eq. (4)], but found to be unsatisfactory, the R2 value being 0.87. A typical fit of the D(t) data to Eq. (4) of the platinum nanocrystals is shown in Fig. 7 by the broken curve. The growth process clearly deviates from diffusion-limited Ostwald ripening. We, therefore, fitted the D(t) data to the surface-limited reaction model (D2 / t) or by varying the value of the exponent (n in Dn / t), and found that the fit of the data was unsatisfactory to both the surface reaction model and to a variable n model. In order to fit the experimental D(t) data of the platinum nanocrystals, we, therefore, used a model which contains both the diffusion-limited and surface-limited growth [2,14,20],

BD3 þ CD2 þ const ¼ t

ð5Þ 1=ðD0 V 2m C 1 Þ,

T=ðkd V 2m C 1 Þ

where B = AT/exp (Ea/kBT), A / c C/ c and kd is the rate constant of surface reaction. We obtained a good fit (R2 = 0.992) over the entire range of experimental data by the combined diffusion and surface reaction control model of Eq. (5) as shown

N. Varghese, C.N.R. Rao / Journal of Colloid and Interface Science 365 (2012) 117–121

by solid curve in Fig. 7. In the presence of PVP, the growth of platinum nanocrystals involves contributions from surface reaction as well. 4. Conclusions The present study shows that diffusion-limited Ostwald ripening satisfactorily explains the growth of platinum nanocrystals in the absence of any capping agent. It is to be noted that platinum nanocrystals formed by the reduction of sodium citrate do not have any strongly binding capping agent. Thus, we find that in the presence of PVP, there is need to take a surface reaction term into account in addition to the Ostwald ripening. Growth of the nanocrystals mainly occurs by the diffusion of monomers from solution to the nanocrystal surface or by the reaction at the surface where the diffusing species get assimilated into the growing nanocrystals. Diffusion and surface reaction are limiting cases in nanocrystal growth. The presence of a capping agent such as PVP gives rise to a barrier to diffusion and the contribution of the surface reaction, therefore, becomes significant. References [1] C.N.R. Rao, P.J. Thomas, G.U. Kulkarni, Nanocrystals: Synthesis, Properties and Applications, Springer, 2007. [2] C.N.R. Rao, A. Muller, A.K. Cheetham (Eds.), Nanomaterials Chemistry: Recent Developments and New Directions, Wiley-VCH, Weinheim, 2007.

[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26]

[27] [28]

121

I.M. Lifshitz, V.V. Slyozov, J. Phys. Chem. Solids 19 (1961) 35. C. Wagner, Z. Elektrochem. 65 (1961) 581. E.M. Wong, J.E. Bonevich, P.C. Searson, J. Phys. Chem. B 102 (1998) 7770. Z. Hu, D.J.E. Ramírez, B.E.H. Cervera, G. Oskam, P.C. Searson, J. Phys. Chem. B 109 (2005) 11209. G. Oskam, A. Nellore, R.L. Penn, P.C. Searson, J. Phys. Chem. B 107 (2003) 1734. F. Mafune, J. Kohno, Y. Takeda, T. Kondow, J. Phys. Chem. B 107 (2003) 4218. S.B. Simonsen, I. Chorkendorff, S. Dahl, M. Skoglundh, J. Sehested, S. Helveg, J. Am. Chem. Soc. 132 (2010) 7968. Z.A. Peng, X. Peng, J. Am. Chem. Soc. 123 (2001) 1389. Z.A. Peng, X. Peng, J. Am. Chem. Soc. 124 (2002) 3343. X. Peng, J. Wickham, A.P. Alivisatos, J. Am. Chem. Soc. 120 (1998) 5343. L. Qu, W.W. Yu, X. Peng, Nano Lett. 4 (2004) 465. K. Biswas, B. Das, C.N.R. Rao, J. Phys. Chem. C 112 (2008) 2404. R. Seshadri, G.N. Subbanna, V. Vijayakrishnan, G.U. Kulkarni, G. Ananthakrishna, C.N.R. Rao, J. Phys. Chem. 99 (1995) 5639. B. Abecassis, F. Testard, O. Spalla, P. Barboux, Nano Lett. 6 (2007) 1723. K. Biswas, N. Varghese, C.N.R. Rao, Small 4 (2008) 649. M.B. Mohamed, Z.L. Wang, M.A. El-Sayed, J. Phys. Chem. A 103 (1999) 10255. D.V. Talapin, A.L. Rogach, M. Haase, H. Weller, J. Phys. Chem. B 105 (2001) 12278. R. Viswanatha, P.K. Santra, C. Dasgupta, D.D. Sarma, Phys. Rev. Lett. 98 (2007) 255501. R. Viswanatha, H. Amenitsch, D.D. Sarma, J. Am. Chem. Soc. 129 (2007) 4470. N. Varghese, K. Biswas, C.N.R. Rao, Chem. Asian. J. 3 (2008) 1435. P.A. Brugger, P. Cuendet, M. Graetzel, J. Am. Chem. Soc. 103 (1981) 2923. T. Herricks, J. Chen, Y. Xia, Nano Lett. 4 (2004) 2367. J.S. Pedersen, Adv. Colloid Interface Sci. 70 (1997) 171. J.S. Pedersen, in: P. Lindner, Th. Zemb (Eds.), Neutrons, X-rays and Light: Scattering Methods Applied to Soft Condensed Matter, North-Holland, New York, 2002, p. 391. J.S. Pedersen, J. Appl. Crystallogr. 37 (2004) 369. L. Rayleigh, Proc. Roy. Soc. London, Ser. A 84 (1911) 25.